And then based on the optimal filtering theory of structure stochastic jump system,a new algorithm for infrared system distance estimation was proposed.

But there are few papers describing modeling-building,optimal filtering algorithms for this kind of systems.

In this paper,an optimal filtering method based on the Singular Value Decomposition (SVD)technique for Multi-Channel Systems with Multiplicative Noise is proposed.

The Optimal Filter and Estimation of a Kind of the Random Signals of the Unmeasurable Components Based on Its Measurable Components in the Generalized Process;

Based on model matching approach,this paper presents an algorithm for designing optimal filter that minimizes the magnitude of the filter error,in the case of linear discrete-time system in which the process noise and measurement noise are unknown but magnitude bounded.

It is important for researchers to find which kinds of filtering equation are solvable and how to give representation of the optimal filter.

Combining Hilbert-Huang Transformation with optimum filtering, because of the narrowness of the spectrum of Intrinsic Mode Functions and guided by the criterion of minimum mean square error,, the necessary and sufficient condition for optimum filtering of random waves was set up through variational calculus.

By use of the matrix singular value decompostion,a practical mothed of designing a reduced-order optimal filter is given.

On the basis of studying fractional Fourier transforms area filtering, fractional Fourier transform optimum filter based least mean-squared error theory is proposed, and is verified by emulation experiment.